Today at Atelier Kaz - ex-Honda R&D, F1, Indy/CART engineer

[ Repost as it didn't show up in the What's New? tab. Origianlly published around 13:50 on 22/Jul/24. ]


As reported earlier, serviced many gauge clusters recently.

I’ll start by stating one thing that has been mentioned so many times.


[ Please do not delay the electronics service if it’s known to fail. ]


From what I saw on this occasion, the level of the damage to the circuit board got even worse than ever.

Some of the circuit boards, gauge dial face, etc are discontinued and no longer available from Honda.

So, the only option is to refurbish/repair the existing ones but it's getting much more time consuming every year.

Even the new one from Honda would fail eventually.

This is because the original circuit board design was from '80s and almost all of the available circuit boards in stock were manufactured well before 2005.

The parts selected at that time are all not RoHs compliant so the manufactures can no longer make new one unless re-evaluate and create new BOM for new circuit board design/spec.

There is a widely known huge issue with many electrolytic capacitors from Japanese manufactures such as Nichicon, Panasonic (Matsu****a), ELM, etc with specific type of fluid contained.

They were in use in '80s and the left over stocked ones were still in use by the third parties until around mid/late '90s.

They will all leak the fluid eventually.


On our NSX, these are some of the components known to fail for the same reason - the capacitor leakage.

A/C Climate Control Unit (CCU)
Gauge Cluster
Bose speaker AMP (3 of them)
Audio Head Unit (Alpine)
SRS control unit (driver side only model)
Headlight adjuster (European model only)

There are additional electronics components that will fail for different reasons (mainly the soldering crack).


Some of the photo from the recent tasks.

This owner contacted me with the intermittent dead speedo.

The needle would stay at 0mph for quite some time but suddenly recovered after driving sometime.

On other occasion, the speedo was working but then started bouncing and eventually dead.

It will be MOT failure unless you can prove you have other means to display accurate enough car speed.

While I couldn't re-create exactly the same issue on the bench, noticed the speedo would 'hang' for short period before reaching the simulator speed.

So, the driver circuit was still operational but most likely contaminated with the leaked acid.






By the way, looking at the dial face, someone already opened this gauge cluster in the past.

On the tacho dial face, almost looked as if something being splashed.






As seen on many gauge clusters, about 300 - 500rpm offset across the mid – high range of tacho.

The offset at the speedo was not easy to tell because of the ‘hang’ effect but enough to spend extra hours for the calibration later.









And this was what I found.

Probably one of the worst leakage so far.

Even some puddle of leaked acid at several places on both circuit boards.

Because many components were soldered with their body so close to the surface of the circuit board, you won't be able to spot the partial short circuit until you measure/test each of them.

I won't go into the detail but found the cause of the 'hang' effect on the speedo.

To clean the area, had to remove many components.

Fortunately, didn't have to remove the HIC for the warning indicator.